Welsh Journals

Search over 450 titles and 1.2 million pages

Roger Turvey, in his detailed study of the Perrot family, concluded that the two probably never existed.9 Tradition further has it that the early castle did not occupy its present site, but was located at Sentence Castle, 2.6 km south of Narberth. Whilst it is agreed that Sentence, a fine example of an earthwork castle, is 12th century in origin, there is no reason to suppose that at any time in history it represented the 'Narberth Castle' referred to in the sources. Furthermore, it has also been suggested that the move to Narberth took place, under the spurious Sir Andrew Perrot, after the Sentence site was destroyed in 1257. The destruction of 'Narberth Castle' in 1257 was a documented event,10 but there is no suggestion in the contemporary chronicles that the castle was relocated. The context of the settlement What is now Narberth lay within the early medieval cwmwd (commote) of Arberth, part of Cantref Penfro which comprised the south Pembroke- shire peninsula, but excluded what are now the parishes of Llanddewi and Lampeter Velfrey. They made up Cwmwd Efelffre, which was part of Cantref Gwarthaf. Cwmwd Arberth appears to have contained a pre-conquest cultural and administrative centre, which appears in no less a work than The Mabinogion. Whilst the location of this centre, with its llys or hall, is unknown, the present castle site is a strong contender. It was common enough for the Anglo- Normans to site their castles at existing administrative centres, while it has been convincingly argued that the route of the A 478, at the foot of the castle spur, pre-dated the Norman conquest (Fig. 1)." The nature of the site suggests that an even earlier, iron age fort cannot be ruled out. The Norman conquest of West Wales was a piecemeal affair conducted by the Anglo-Norman aristocracy under their own initiative. Having estab- lished a base at Pembroke with his father Roger, in 1093, the Norman Amulf de Montgomery set about the task of subduing the surrounding area. It must be stressed that during these initial few years no formal government of these conquests was in place, and no boundaries drawn. Neither the extent of Amulf's territories, nor the nature of his tenure during this early period is known, but by 1102 a large slice of modern